Hindsight Bias Still Dominates Technical Analysis

Hindsight bias, or the tendency to see past events as predictable, still dominates technical analysis even though it has become much easier to check the validity of any claims. I present an example from the healthcare sector and then I attempt to identify the reasons of this widespread phenomenon of post hoc analysis.

When a security reaches all-time highs, there are usually a number of articles in the financial media and blogosphere that attempt to justify price action based of post hoc use of technical analysis. Although such analysis is the result of a serious form of bias called hindsight, nevertheless some of the analysts that resort to it, knowingly or unknowingly, manage to impress crowds.

Let us consider the health care ETF (XLV) example below.

As it may be seen from the top chart pane above, XLV made a new all-time high last week. There are numerous articles from aggressive analysts that attempt to intimidate investors by pointing to the above chart and to missed profits by those that did not follow their post hoc analysis. Obviously, none of these analysts had forecasted the uptrend a few years back, ex-ante.

The second pane of the above chart shows the buy and hold performance since inception of the ETF. Annualized return (CAGR) is 8.02% at 39.17% maximum drawdown, resulting in a MAR ratio of 0.20 (CAGR/Max. DD.)

However, a long-only 50/200 golden cross strategy would have realized less than half of the above annualized return, only 3.44%, since inception due to price whipsaw, as it may be seen in the third pane. MAR for the strategy is even much lower at 0.08 due to higher maximum drawdown.

If fact, there is no easily conceived strategy that would have outperformed buy and hold in XLV. Our proprietary mean-reversion strategy, known as PSI5 (MR5), has managed to match the risk-adjusted MAR performance but underperforms buy and hold on an annualized return basis, as shown below:

The conclusion is that there has been no easy ex-ante strategy for capturing XLV returns. Any analysis driven by references to all-time highs is hindsight bias. But why do so many use hindsight bias in their analysis? Below is a list of possible reasons.

  1. They know no better way from this mainstream approach to technical analysis
  2. They lack quantitative skills to access their claims
  3. They know many get impressed by hindsight analysis
  4. Many hesitate to blast hindsight analysis from fear of being called haters
  5. Many who use hindsight bias are high profile analysts at investment firms
  6. Hindsight is easy, forecasting is hard
  7. Technical analysis works perfectly in hindsight mode

Paying attention to analysis based on hindsight and to those that employ it can lead to nasty surprises and losses. But in books and blogs, only the winning cases will be displayed usually with an extra dose of arrogance.

If you found this article interesting, I invite you follow this blog via any of these methods: RSS or Email, or follow us on Twitter

If you have any questions or comments, happy to connect on Twitter: @mikeharrisNY

Charting and backtesting program: Amibroker

Disclaimer

Technical and quantitative analysis of Dow-30 stocks and 30 popular ETFs is included in our Weekly Premium Report. Market signals for longer-term traders are offered by our premium Market Signals service. Mean-reversion signals for short-term SPY traders are provided in our Mean Reversion report.

Copyright Notice

This entry was posted in Technical Analysis and tagged , . Bookmark the permalink.

Leave a Reply